[EECBandit]

I have a problem where under extremely light throttle to coasting, the ecu pulls a lot of fuel. As soon as I give it some gas the KAMRF's bounce up to within 1-2% of 1.

I have tried cleaning up the MAF below the Idle MAFv, but decel still pulls almost all KAMRF 11%.

Is there a way to cut some decel fuel so that the ECU doesn't try to pull as much fuel?


Running a 331 with a A9P EEC.

Thanks.

[skunk]

Have you logged Throttle Status to see if your "light throttle* is registered as a closed throttle position? Also look at ISCDC .....it's likely in daspot mode. You could experiment with pre-position or bump FKARCH up a hair but that would also fatten up everything down low and require minor adjustments to your MAF curve. You could also reduce the lean HEGO Bias down low by moving the numbers toward negative (rich)

You have to consider that at low loads your Kamrf is learning from what supposed to be your idle cells. If your idle is increased above stock you may need to alter the adaptive learning table to reflect the idle cell.

Personally I wouldn't worry a whole lot about a lean decel. It actually is easier for the engine to recover on tip-in. I can't remember stock characteristics since I've enable DFSO based on the X3Z.

John

[EECBandit]

Thank John,

I like the idea of not worrying about it. The car runs fine, just trying to see if I can tweak it a bit. I was looking at MEFTRD and might give that a decrease and see how it responds.

[skunk]

Just remember that is part of the Transient Fuel system and isn't in the mix for long.

The more I think about it.....the more I think you need to take a look at the adaptive learning table to reflect properly. I am basing my opinion on you comment about KAMRF variance. This is the only place that governs learned fuel.

I don't know what your Idle RPM is but its highly doubtful it's stock. Look at FN1325 loads below 20 all reference cell-80 which is 625 RPM . If your closer to 900 RPM change all the -80 to -82. That should force decel trims to reflect closed throttle idle trims.

It won't effect anything else so if you don't like the outcome, just flip it back to how you had it.

John

[EECBandit]

Thanks John, that's a great suggestion. I have 4 throughout the table right now but never changed RPM scale. I have my idle set at 824 so I am going to try and reference another cell for below idle and low load.

[skunk]

Well there's the answer. Setting closed throttle and decel to learn is going to bounce KAMRF around alot. You don't have to re-scale rpm to correct the issue but its nice to have it reflect the range your using.

If you employ that strategy of fast learning leave all -80 alone or set to reference your idle RPM. If your at 824 you could go with either -81 or -82. Then go-to Z_LTMT880 thru 883 and set that as desired.

John

[EECBandit]

I'll definitely give it a try and report back. won't be able to do it for a few days because of work and other commitments.

Wish I could try it sooner :)

[EDS50]

Is DFSO or MFA enabled?

[EECBandit]

DSFO Z_DFSW is enabled.

Haven't had a chance to play around with my tune as it's been raining.

I was also considering to reset all parameter back to default A9P to see if I've screwed something else up without knowing.

Obviously the the critical items like MAF, fuel injectors param and timing tables, etc will stay where they are at.

[EDS50]

Disable dfso and mfa and your decel conditions wont be so drastically lean as you are reporting. Also pay attention to the maf xfer region where you are at during decel and adjust accordingly.

[EECBandit]

If I reduce the MAF around the coast MAFV, then the KAMRF will add fuel at idle. I think I have the MAF as clean as possible.

I'll try to disable DFSO and MFA and report back.

Thanks for the ideas....

[EDS50]

Keep an eye on the maf xfer with the table trace function as it is possible for the mafv to dip below idke mfv cells during decel conditions. Also reducing kg/hr will lean out your afr

[EECBandit]

Well, I finally had some time to look at my tune.
While checking DFSO, I noticed that DSFRPM Fuel - DFSO Minimum RPM = 6000

Doesn't this basically disable DFSO even if the switch is on?

[EDS50]

Well, I finally had some time to look at my tune.
While checking DFSO, I noticed that DSFRPM Fuel - DFSO Minimum RPM = 6000

Doesn't this basically disable DFSO even if the switch is on?

DSFRPM - The minimum rpm above the desired idle rpm at which deceleration fuel shut off is permitted.RPM above DSDRPM (Idle speed control desired RPM) where Decel Fuel Shutoff is still active

N - DSDRPM > DSFRPM - Decel fuel shut-off is still active
N - DSDRPM < DSFRPM - DSFRPH - Decel fuel shut-off is turned off

Example:
DSDRPM = 780
DSFRPM = 1050 (stock setting)
DSFRPH = 100 (stock setting)

Above RPM limit -
RPM = 2000
2000 - 780 = 1220 > 1050 so decel fuel shut-off is still active

Within hysteresis window -
RPM = 1800
1800 - 780 = 1020 < 1050 but;
1800 - 780 = 1020 > 950 (1050 - 100) so decel fuel shut-off is still active

Below threshold -
RPM = 1700
1700 - 780 = 920 < 1050 and;
1700 - 780 = 920 < 950 (1050 - 100) - decel fuel shut-off is turned off

[EECBandit]

Any idea why the base cal would have 6000 for a default DSFRPM value?

Now I'm worried that if I turn on or correct DFSO I may end up opening a whole new can of worms

[EDS50]

It appears to me the notes and the setting vary tremendously. Ill look into it more.

[skunk]

Any idea why the base cal would have 6000 for a default DSFRPM value?

Now I'm worried that if I turn on or correct DFSO I may end up opening a whole new can of worms

I read through very quickly but 6000 is not correct As Ed points out...stock is 1050, I have no idea if the A9P is different. If you look at the X3Z....I believe it's 150.

Either way....you have it disabled, I would still turn it off via the switch.

John

[cgrey8]

I haven't confirmed this by actually looking, but I do recall that A9L had DFSO, technically enabled, but effectively disabled given the allowed RPM was higher than anybody would sustain. And I believe it was that minimum RPM scalar that basically disabled it. The X3Z, however had DFSO enabled and it had a minimum RPM that would allow the code to actually shut off the injectors.

Back 15 years ago this was an important distinction because our GUFx definitions weren't as broken out as they are today. So people (myself included) started with stock X3Z.bin files to get DFSO functional in their base tunes. Once the defs got more mature and the necessary DFSO scalars were exposed, the importance and desire to use the X3Z as a base was no longer as relevant. In fact, I think the A9L was generally deemed a better base tune than the X3Z.

[EDS50]

Well, I finally had some time to look at my tune.
While checking DFSO, I noticed that DSFRPM Fuel - DFSO Minimum RPM = 6000

Doesn't this basically disable DFSO even if the switch is on?

DSFRPM - The minimum rpm above the desired idle rpm at which deceleration fuel shut off is permitted.RPM above DSDRPM (Idle speed control desired RPM) where Decel Fuel Shutoff is still active

N - DSDRPM > DSFRPM - Decel fuel shut-off is still active
N - DSDRPM < DSFRPM - DSFRPH - Decel fuel shut-off is turned off

Example:
DSDRPM = 780
DSFRPM = 1050 (stock setting)
DSFRPH = 100 (stock setting)

Above RPM limit -
RPM = 2000
2000 - 780 = 1220 > 1050 so decel fuel shut-off is still active

Within hysteresis window -
RPM = 1800
1800 - 780 = 1020 < 1050 but;
1800 - 780 = 1020 > 950 (1050 - 100) so decel fuel shut-off is still active

Below threshold -
RPM = 1700
1700 - 780 = 920 < 1050 and;
1700 - 780 = 920 < 950 (1050 - 100) - decel fuel shut-off is turned off

The notes I posted above are the from the pid descriptor for DSFRPM from the GUF1 strategy. GUFB reads as:

DSFRPM - The minimum rpm above the desired idle rpm at which deceleration fuel shut off is permitted. RPM above DSDRPM (Idle speed control desired RPM) where Decel Fuel Shutoff is still active

N - DSDRPM > DSFRPM - Decel fuel shut-off is still active
N - DSDRPM < DSFRPM - DSFRPH - Decel fuel shut-off is turned off

Example:
DSDRPM = 780
DSFRPM = 1050 (stock setting)
DSFRPH = 100 (stock setting)

Above RPM limit -
RPM = 2000
2000 - 780 = 1220 > 1050 so decel fuel shut-off is still active

Within hysteresis window -
RPM = 1800
1800 - 780 = 1020 < 1050 but;
1800 - 780 = 1020 > 950 (1050 - 100) so decel fuel shut-off is still active

Below threshold -
RPM = 1700
1700 - 780 = 920 < 1050 and;
1700 - 780 = 920 < 950 (1050 - 100) - decel fuel shut-off is turned off

The GUFB.doc reads as follows:

FUEL STRATEGY - DECEL FUEL SHUT OFF - GUE0
PEDD-PTOPE, FoMoCo, PROPRIETARY & CONFIDENTIAL
DECEL FUEL SHUT OFF STRATEGY

The Fuel Shutoff strategy is divided into two sub-strategies:

1) Decel Fuel Lean Out.
2) Manual Transmission Shift Fuel Lean Out.
The Decel Fuel Lean Out strategy reduces the fuel flow during specific load
and RPM condition only during Closed Throttle Mode. The FUELPWs are
multiplied by FN374 during a decel until the RPM is within the band created
by [DSFRPM - DSFRPH] of the idle RPM range. Typically, FN374 is calibrated
to provide a lean limit air/fuel ratio during the decel RPM range.
The Manual Transmission Shift Fuel Lean out strategy reduces (or eliminates)
fuel flow at the beginning of the shift, until the RPM is within the band
created by [SHFRPM - SHFHYS] of the idle speed (or until the transmission has
been in Neutral for DSTM2 seconds. Typically, FN374 can be calibrated to
zero during the range of shift RPMs.




CALIBRATION HINTS:
The use of Decel Fuel Shutoff can aggravate Clunk. Therefore, DFSO should occur only
at low airflow (low MAP) in order to minimize the rate of change of Torque. CTDFSO can
be used to delay DFSO.
FN374 should be calibrated to avoid lean misfires. Therefore, it should be
either zero (Decel Fuel Shutoff) or greater than the lean limit (Decel Fuel
Lean Out).
CTEDSO should be greater than CTDFSO to prevent DFSO after Tip-in, Tip-out.
For MANUAL TRANSMISSIONS: SHFRPM should be greater than DSFRPM.
6-73

Now, I do realize the GUFB.doc notes and the notes I see in the pid descriptors are meant for a t-5 trans and not aod. I have also come to realize that the document notes, pid descriptors and actual default .bin setting for a9p and a9l (and x3z is completely different) vary and do not represent the pid descriptors or strategy so it looks like we need to dig deeper into this to decipher what is right and how the default .bins should be set. I don't have a GUF1.doc to compare to the GUFB.doc.

[EECBandit]

I turned off DFSO via the switch, went out for a 30 mile ride and no difference in the decel. Not sure it worth spending a whole lot of time on it as it's only during decel, it seems to run well otherwise.

I play around with the tune more over the weekend if the weather cooperates.

Really appreciate the help.

[EDS50]

Try to disable mfasw and clear your kamrfs and report back. There is an easy fix to your issue as long as we explore all options.

[EDS50]

Please post your current tune also so we can look into things a little deeper.

[EECBandit]

Tune Attached.

[EDS50]

Looking at your tune I am assuming you are using 42# injectors. Where did you get your injector settings from? Can you post a datalog, I am curious to see if under decal you are below 0.500 mafv. Your breakpoint is way out in left field for your injectors and I would most definitely disable mfa. (mfasw)

[EECBandit]

I just disabled MFA in my tune.

I am using 34 / 36 lb ford racing injectors.
Unfortunately I don't have a small data log, they are all huge. I will try and get a reasonable size log this afternoon.
I've attached the data sheet for my injectors

[EDS50]

That makes more sense then. Are you using a calibrated maf? You can look in your logs to see if mafv dips below 0.500 under decal conditions. Try disabling mfasw first and reset your kamrfs and see if that helps your lean decel.

[EECBandit]

I just disabled disabled mfasw in my tune and will reset kamrfs.


I have a 90mm LMAF that I built / tweaked the curve from the starting MAF transfer I got from EFIDyno.

[EDS50]

Ok sounds good. Lets us know how things go once the new kamrfs rebuild. Keep an eye on mafv under decal conditions to see if it dips below 0.500 under decal conditions. Its an easy fix if it does.

[EECBandit]

Will do.

Thanks for your help, really appreciate it.

[EDS50]

No problem. Anytime.